The amino acid sequence of a protein determines its secondary, tertiary, and quaternary structure to result in the protein's final three-dimensional (3D) shape. The shape and functional groups (side chains) of the amino acids therein define the protein's function. In the case of a binding protein, the portion of the protein responsible for the binding activity (binding domain) must either be exposed, or be capable of being exposed, on an accessible surface of the protein exposed to the exterior solvent to provide for possible interaction with a binding target. Thus to vary the binding activity, the amino acid residues of the binding domain must be varied.
With an immunoglobulin as an example of a familiar binding protein with specificity and affinity, the “variable region” or binding domain includes six loops clustered in space. The loops provide the 6 complementarity determining regions (CDRS) and are contained in two polypeptides, a heavy chain and a light chain, each carrying 3 CDRs (H1, H2, and H3 of the heavy chain and L1, L2, and L3 of the light chain). The amino acid residues of the variable regions orient the CDRs toward the exterior solvent environment to permit their interaction with an antigen. High sequence variability of the amino acid residues of the CDRs allows immunoglobulins as a class to bind a large variety of antigens. The CDRs and non-CDR portion of the variable region form an immunoglobulin fold to determine the structure of the loops and thereby maintain the overall structure of the immunoglobulin variable region, with proper orientation of the CDRs.
But variability in the sequence of a protein, like an immunoglobulin, is often limited by the effects of variability on protein folding and the resulting final 3D shape. Amino acid residues with side chains that are not exposed to the exterior solvent are often limited in variability because as part of the protein's interior they must “fit” within the interior space as dictated by other amino acid residues. The protein can tolerate greater variability in residues with side chains oriented toward, and exposed to, the exterior solvent, given that they do not have to “fit” into an interior space constrained by other residues.
To diversify the binding functionality of a binding protein and thus promote recognition of members of a diverse population of target molecules, amino acid variability is necessary. Interactions between a binding protein and its target molecule (the ligand) are usually non-covalent and yet often very tight (high affinity or avidity) and specific. The intermolecular interactions are defined by the amino acid residues of the protein's binding domain which form a surface that fits “hand-in-glove” like onto the surface of the ligand being bound. The two contacting surfaces must have complementarity via hydrogen bonding (at times mediated by a water molecule), charge interactions, alignment of attracting dipoles, hydrophobic to hydrophobic (van der Waals) interactions, and/or protrusions fitting with depressions.
In the example of an immunoglobulin, the binding domain is presented within the context of the framework made up by the rest of the immunoglobulin molecule. The framework, generally referred to as the immunoglobulin fold, forms the scaffold of the protein structure and functions to correctly present the binding domain. The framework restrains the 3D shape of the protein so that the amino acid residues of the binding domain are positioned in a manner to create the accessible specific binding site.
The usefulness of immunoglobulins as manipulable binding proteins is limited, however, by the nature of the immunoglobulin framework, which requires two polypeptides to form the complete ligand- or antigen-binding site. This results in a number of disadvantages: the need to manipulate rather large polypeptides, the need for complicated molecular cloning to diversify a binding site; and the complication of modifying six different CDRs. The consequences of these disadvantages include constraints on using phage display (see for example U.S. Pat. Nos. 5,223,409 and 5,571,698) to diversify immunoglobulins for the purpose of creating new binding or other functional activities.
A number of attempts have been made to overcome the limitations of immunoglobulins. These include the use of a CTL4-like sandwich architecture as a framework for presenting randomized peptide sequences (see WO 00/60070); the use of fibronectin type III domains (see U.S. Pat. No. 6,818,418); the use of an “anticalin” (see WO 99/16873 and Beste et al. Proc. Natl. Acad. Sci., USA 96:1898-1903 (1999)); and even the use of single chain antibodies, optionally with a CH3 domain of an immunoglobulin to permit spontaneous dimerization.
Citation of documents herein is not intended as an admission that any is pertinent prior art. All statements as to the date or representation as to the contents of documents is based on the information available to the applicant and does not constitute any admission as to the correctness of the dates or contents of the documents.